0616
Complex B1+ field-based conductivity mapping in the human myocardium at 3T1TU Delft, Delft, Netherlands, 2UMC Utrecht, Utrecht, Netherlands, 3Linköping University, Linköping, Sweden, 4Holland PTC, Delft, Netherlands
Synopsis
Keywords: Myocardium, Electromagnetic Tissue Properties, electrical properties, conductivity, parametric mapping, cardiac, EPT
Motivation: While electrical property tomography is gaining popularity, cardiac applications are limited due to inadequate cardiac B1+ mapping. Thus, conductivity mapping in the heart using complex B1+ maps is yet unexplored.
Goal(s): To measure myocardial conductivity from the complex B1+ distribution in the heart at 3T.
Approach: A novel |B1+| mapping method was adapted for free-breathing B1+ maps, followed by conductivity reconstruction via 1D polynomial fitting in saline phantoms and four healthy subjects.
Results: Phantom results show excellent correlation with expected values (R2=0.95). In-vivo, conductivity is largely homogenous with 0.69±0.13S/m average in-plane conductivity across all subjects, in line with the literature.
Impact: Electrical properties are a valuable biomarker, however, the translation to cardiac imaging remains limited. In this work, complex B1+ field-based conductivity is reported in the human myocardium at 3T, using a novel Bloch-Siegert shift-prepared cardiac B1+ mapping technique.
Introduction
While electrical property tomography (EPT) has gained attention in neuro and breast imaging, its translation to the heart is greatly limited [1]. High inherent sensitivity to residual motion and noise of the EPT reconstruction render poor B1+ mapping quality a major limitation. Previous efforts to map conductivity in the heart have relied on transceive phase maps, albeit requiring limiting assumption of constant B1+ magnitude [2]. In this work, we aim to reconstruct myocardial conductivity by adapting a recently proposed free-breathing B1+ mapping technique to obtain complex B1+ magnitude and phase maps.Methods
B1+ field magnitude and phase maps were obtained using a novel Bloch-Siegert (BS) shift-prepared cardiac B1+ mapping technique [3]. The sequence acquires three baseline images for the magnitude maps (Fig. 1a,b): one with BS preparation inducing 4φBS phase shift (IBS1); one with equivalent preparation leading to zero phase shift (IBS2), for magnetization transfer effect compensation; and a saturation-prepared image for capturing image readout (ISAT). To increase the SNR in magnitude and phase B1+ maps, nBS repetitions were obtained for IBS1/IBS2. Finally, to obtain additional phase data, nP extra non-prepared images were acquired. The B1+ transceive phase was reconstructed as the average of PBS1 and PP phase images. B1+ phase was assumed to be equal to half of the measured transceive phase [4] (Fig. 1c).All imaging was performed at 3T (Ingenia, Philips). The B1+ maps were acquired in a custom-made phantom (coronal and transverse plane), with 9 vials containing different saline solutions. In-vivo data was acquired in the four-chamber view for 4 healthy subjects (3 male, age 29±3). B1+ mapping was performed with ECG-triggering (end-diastole) and during free-breathing using a pencil beam diaphragmatic navigator. Single-shot bSSFP readout with 40° flip angle was used (TR/TE = 2.4/1.2ms, 1.9x1.9x10/300x300x10mm3 resolution/FOV). The BS preparation block comprised a Fermi pulse with 4.5ms duration and 7kHz off-resonance. nBS=20/nBS=19 baseline preparations were used in phantom/in-vivo, with nP=9 additional images. Rest periods of 5s were used after each BS-prepared image for magnetization recovery resulting in a total scan time of 4 minutes (assuming 100% navigator efficiency). Groupwise registration was applied to the real and imaginary part of the acquired images to minimize residual motion [5].
Conductivity (σ) was reconstructed using a 1D polynomial fitting approach [6] on the complex-valued B1+ data (Fig. 2). For the phantom, 1D reconstructions were done in all three directions and subsequently summed. In-vivo data region of interest (ROI) was manually drawn in the septum, excluding voxels affected by partial volume effects. This was followed by polynomial 1D conductivity fitting along the longest direction of the mask, multiplied by 2 for in-plane conductivity. Equal contributions from the two in-plane directions were assumed to prevent noisy contributions from the short-axis direction.
Results
Masked complex B1+ component maps are illustrated in Fig. 3a. A consistent field curvature is clearly visible in the phase images, and, to a lesser extent in the magnitude maps. Phantom conductivity maps (Fig. 3b) show a low in-vial variation (<0.06 S/m) and clear contrast between the different saline concentrations. Excellent correlation (R2=0.95) with the values expected from the prepared salinity (0.4-1.2 S/m) is observed [7].Representative subject maps are shown in Fig. 4a. The B1+ magnitude and phase in the septum (masked region) exhibit a gradient along the long axis. Reconstructed in-plane (2x1D) conductivity σ’ (0.74±0.02 S/m) is homogeneous throughout the ROI. The reconstructed σ’ values showed an overall mean of 0.69±0.13 S/m across all subjects (Fig. 5).
Discussion
In this work, we present a new method for robust cardiac conductivity mapping using B1+ magnitude and phase extracted from a BS-prepared cardiac B1+ mapping sequence. In-vivo, an excellent agreement was achieved with the previously measured phase-based conductivity of 0.69 S/m [4].Only conductivity was extracted from the acquired complex B1+ maps. Additional permittivity reconstruction is also feasible, however, it is more sensitive to noise in B1+ magnitude. To provide sufficient |B1+| quality, an increased number of baseline images can be acquired with the proposed free-breathing mapping technique by scaling the scan time. Thus, its application for robust cardiac conductivity and permittivity imaging warrants further investigation.
The proposed technique was acquired in 2D, thus, requiring boundary assumptions in the electrical property reconstruction. However, as the B1+ preparation is independent from the readout, the acquisition can easily be scaled to volumetric imaging. This warrants an exploration for more accurate cardiac EPT without the need for spatial assumptions.
Conclusions
The proposed complex B1+ mapping enables robust reconstruction of conductivity maps in the human heart from a single free-breathing acquisition.Acknowledgements
S.M. acknowledges funding from the Netherlands Organization for Scientific Research (NWO; VENI grant no. 18078). S.W. acknowledges funding from the NWO (Start-up STU.019.024), and the European Union (ERC, VascularID, StG 101078711).References
[1] Katscher U, and van den Berg CAT. Electric properties tomography: Biochemical, physical and technical background, evaluation and clinical applications. NMR in Biomedicine 2017; 30.8: e3729.
[2] Katscher, U, Weiss, S. Mapping electric bulk conductivity in the human heart. Magn Reson Med. 2022; 87: 1500–1506.
[3] Šiurytė P, Henningsson M, van de Steeg-Henzen C, Weingärtner S. |B1+|-prepared imaging for efficient cardiac transmit field mapping. ISMRM 31st Annual Meeting, 3-8 June 2023
[4] Van Lier AL, Raaijmakers A, Voigt T, Lagendijk JJ, Luijten PR, Katscher U, van den Berg, C A. Electrical properties tomography in the human brain at 1.5, 3, and 7T: a comparison study. Magn Reson Med. 2014. 71(1), 354-363.
[5] Tao Q, van der Tol P, Berendsen FF, Paiman EHM, Lamb HJ and van der Geest RJ. Robust motion correction for myocardial T1 and extracellular volume mapping by principle component analysis-based groupwise image registration. J. Magn. Reson. Imaging 2018; 47: 1397-1405.
[6] Karsa A, Shmueli K. New Approaches for Simultaneous Noise Suppression and Edge Preservation to Achieve Accurate Quantitative Conductivity Mapping in Noisy Images. ISMRM 29th Annual Meeting, 15-20 May 2021.
[7] Stogryn A. Equations for Calculating the Dielectric Constant of Saline Water (Correspondence). IEEE Transactions on Microwave Theory and Techniques. 1971. vol. 19, no. 8, pp. 733-736.
Figures
